Skip to content

L
linux
Commit 318103c4 authored by Jay Vosburgh's avatar Jay Vosburgh Committed by Dave Jones
Browse files
Options

[bonding] bug fixes, and a few minor feature additions 

Mainly sync w/ 2.4.x version.
parent 032a8560
Show whitespace changes
Inline Side-by-side
Showing with 461 additions and 269 deletions
Documentation/networking/bonding.txt
View file @ 318103c4
......@@ -7,6 +7,7 @@ Corrections, HA extensions : 2000/10/03-15 :
- Constantine Gavrilov <const-g at xpert.com>
- Chad N. Tindel <ctindel at ieee dot org>
- Janice Girouard <girouard at us dot ibm dot com>
- Jay Vosburgh <fubar at us dot ibm dot com>
Note :
------
......@@ -199,28 +200,42 @@ It is critical that either the miimon or arp_interval and arp_ip_target
parameters be specified, otherwise serious network degradation will occur
during link failures.
mode
Specifies one of four bonding policies. The default is round-robin.
Possible values are:
0 Round-robin policy: Transmit in a sequential order from the
first available slave through the last. This mode provides
load balancing and fault tolerance.
max_bonds
1 Active-backup policy: Only one slave in the bond is active. A
different slave becomes active if, and only if, the active slave
fails. The bond's MAC address is externally visible on only
one port (network adapter) to avoid confusing the switch.
This mode provides fault tolerance.
Specifies the number of bonding devices to create for this
instance of the bonding driver. E.g., if max_bonds is 3, and
the bonding driver is not already loaded, then bond0, bond1
and bond2 will be created. The default value is 1.
mode
2 XOR policy: Transmit based on [(source MAC address XOR'd with
destination MAC address) modula slave count]. This selects the
same slave for each destination MAC address. This mode provides
load balancing and fault tolerance.
3 Broadcast policy: transmits everything on all slave interfaces.
This mode provides fault tolerance.
Specifies one of four bonding policies. The default is
round-robin (balance-rr). Possible values are (you can use either the
text or numeric option):
balance-rr or 0
Round-robin policy: Transmit in a sequential order
from the first available slave through the last. This
mode provides load balancing and fault tolerance.
active-backup or 1
Active-backup policy: Only one slave in the bond is
active. A different slave becomes active if, and only
if, the active slave fails. The bond's MAC address is
externally visible on only one port (network adapter)
to avoid confusing the switch. This mode provides
fault tolerance.
balance-xor or 2
XOR policy: Transmit based on [(source MAC address
XOR'd with destination MAC address) modula slave
count]. This selects the same slave for each
destination MAC address. This mode provides load
balancing and fault tolerance.
broadcast or 3
Broadcast policy: transmits everything on all slave
interfaces. This mode provides fault tolerance.
miimon
......@@ -229,6 +244,27 @@ miimon
100 is a good starting point. See High Availability section for
additional information. The default value is 0.
use_carrier
Specifies whether or not miimon should use MII or ETHTOOL
ioctls vs. netif_carrier_ok() to determine the link status.
The MII or ETHTOOL ioctls are less efficient and utilize a
deprecated calling sequence within the kernel. The
netif_carrier_ok() relies on the device driver to maintain its
state with netif_carrier_on/off; at this writing, most, but
not all, device drivers support this facility.
If bonding insists that the link is up when it should not be,
it may be that your network device driver does not support
netif_carrier_on/off. This is because the default state for
netif_carrier is "carrier on." In this case, disabling
use_carrier will cause bonding to revert to the MII / ETHTOOL
ioctl method to determine the link state.
A value of 1 enables the use of netif_carrier_ok(), a value of
0 will use the deprecated MII / ETHTOOL ioctls. The default
value is 1.
downdelay
Specifies the delay time in milli-seconds to disable a link after a
......@@ -277,14 +313,17 @@ primary
multicast
Integer value for the mode of operation for multicast support.
Option specifying the mode of operation for multicast support.
Possible values are:
0 Disabled (no multicast support)
disabled or 0
Disabled (no multicast support)
1 Enabled on active slave only, useful in active-backup mode
active or 1
Enabled on active slave only, useful in active-backup mode
2 Enabled on all slaves, this is the default
all or 2
Enabled on all slaves, this is the default
Configuring Multiple Bonds
......@@ -321,7 +360,52 @@ For just a single target the options would resemble:
alias bond0 bonding
options bond0 arp_interval=60 arp_ip_target=192.168.0.100
Potential Problems When Using ARP Monitor
=========================================
1. Driver support
The ARP monitor relies on the network device driver to maintain two
statistics: the last receive time (dev->last_rx), and the last
transmit time (dev->trans_start). If the network device driver does
not update one or both of these, then the typical result will be that,
upon startup, all links in the bond will immediately be declared down,
and remain that way. A network monitoring tool (tcpdump, e.g.) will
show ARP requests and replies being sent and received on the bonding
device.
The possible resolutions for this are to (a) fix the device driver, or
(b) discontinue the ARP monitor (using miimon as an alternative, for
example).
2. Adventures in Routing
When bonding is set up with the ARP monitor, it is important that the
slave devices not have routes that supercede routes of the master (or,
generally, not have routes at all). For example, suppose the bonding
device bond0 has two slaves, eth0 and eth1, and the routing table is
as follows:
Kernel IP routing table
Destination Gateway Genmask Flags MSS Window irtt Iface
10.0.0.0 0.0.0.0 255.255.0.0 U 40 0 0 eth0
10.0.0.0 0.0.0.0 255.255.0.0 U 40 0 0 eth1
10.0.0.0 0.0.0.0 255.255.0.0 U 40 0 0 bond0
127.0.0.0 0.0.0.0 255.0.0.0 U 40 0 0 lo
In this case, the ARP monitor (and ARP itself) may become confused,
because ARP requests will be sent on one interface (bond0), but the
corresponding reply will arrive on a different interface (eth0). This
reply looks to ARP as an unsolicited ARP reply (because ARP matches
replies on an interface basis), and is discarded. This will likely
still update the receive/transmit times in the driver, but will lose
packets.
The resolution here is simply to insure that slaves do not have routes
of their own, and if for some reason they must, those routes do not
supercede routes of their master. This should generally be the case,
but unusual configurations or errant manual or automatic static route
additions may cause trouble.
Switch Configuration
====================
......@@ -462,7 +546,7 @@ Frequently Asked Questions
If not explicitly configured with ifconfig, the MAC address of the
bonding device is taken from its first slave device. This MAC address
is then passed to all following slaves and remains persistent (even if
the first slave is removed) until the bonding device is brought
the the first slave is removed) until the bonding device is brought
down or reconfigured.
If you wish to change the MAC address, you can set it with ifconfig:
......@@ -606,12 +690,16 @@ backup, use ifconfig. All backup interfaces have the NOARP flag set.
To use this mode, pass "mode=1" to the module at load time :
# modprobe bonding miimon=100 mode=active-backup
or:
# modprobe bonding miimon=100 mode=1
Or, put in your /etc/modules.conf :
alias bond0 bonding
options bond0 miimon=100 mode=1
options bond0 miimon=100 mode=active-backup
Example 1: Using multiple host and multiple switches to build a "no single
point of failure" solution.
......@@ -698,7 +786,7 @@ allows to reduce down-time if the value of updelay has been overestimated.
Examples :
# modprobe bonding miimon=100 mode=1 downdelay=2000 updelay=5000
# modprobe bonding miimon=100 mode=0 downdelay=0 updelay=5000
# modprobe bonding miimon=100 mode=balance-rr downdelay=0 updelay=5000
Promiscuous Sniffing notes
......
drivers/net/bonding.c
View file @ 318103c4
......@@ -258,10 +258,47 @@
* - correct ifr_data reference. Update ifr_data reference
* to mii_ioctl_data struct values to avoid confusion.
*
*
* 2002/11/22 - Bert Barbe <bert.barbe at oracle dot com>
* - Add support for multiple arp_ip_target
*
* 2002/12/13 - Jay Vosburgh <fubar at us dot ibm dot com>
* - Changed to allow text strings for mode and multicast, e.g.,
* insmod bonding mode=active-backup. The numbers still work.
* One change: an invalid choice will cause module load failure,
* rather than the previous behavior of just picking one.
* - Minor cleanups; got rid of dup ctype stuff, atoi function
*
* 2003/02/07 - Jay Vosburgh <fubar at us dot ibm dot com>
* - Added use_carrier module parameter that causes miimon to
* use netif_carrier_ok() test instead of MII/ETHTOOL ioctls.
* - Minor cleanups; consolidated ioctl calls to one function.
*
* 2003/02/07 - Tony Cureington <tony.cureington * hp_com>
* - Fix bond_mii_monitor() logic error that could result in
* bonding round-robin mode ignoring links after failover/recovery
*
* 2003/03/17 - Jay Vosburgh <fubar at us dot ibm dot com>
* - kmalloc fix (GFP_KERNEL to GFP_ATOMIC) reported by
* Shmulik dot Hen at intel.com.
* - Based on discussion on mailing list, changed use of
* update_slave_cnt(), created wrapper functions for adding/removing
* slaves, changed bond_xmit_xor() to check slave_cnt instead of
* checking slave and slave->dev (which only worked by accident).
* - Misc code cleanup: get arp_send() prototype from header file,
* add max_bonds to bonding.txt.
*
* 2003/03/18 - Tsippy Mendelson <tsippy.mendelson at intel dot com> and
* Shmulik Hen <shmulik.hen at intel dot com>
* - Make sure only bond_attach_slave() and bond_detach_slave() can
* manipulate the slave list, including slave_cnt, even when in
* bond_release_all().
* - Fixed hang in bond_release() while traffic is running.
* netdev_set_master() must not be called from within the bond lock.
*
* 2003/03/18 - Tsippy Mendelson <tsippy.mendelson at intel dot com> and
* Shmulik Hen <shmulik.hen at intel dot com>
* - Fixed hang in bond_enslave(): netdev_set_master() must not be
* called from within the bond lock while traffic is running.
*/
#include <linux/config.h>
......@@ -278,6 +315,7 @@
#include <linux/init.h>
#include <linux/timer.h>
#include <linux/socket.h>
#include <linux/ctype.h>
#include <linux/errno.h>
#include <linux/netdevice.h>
#include <linux/inetdevice.h>
......@@ -288,7 +326,7 @@
#include <linux/if_bonding.h>
#include <linux/smp.h>
#include <linux/if_ether.h>
#include <linux/if_arp.h>
#include <net/arp.h>
#include <linux/mii.h>
#include <linux/ethtool.h>
......@@ -298,8 +336,8 @@
#include <asm/dma.h>
#include <asm/uaccess.h>
#define DRV_VERSION "2.4.20-20021210"
#define DRV_RELDATE "December 10, 2002"
#define DRV_VERSION "2.5.65-20030320"
#define DRV_RELDATE "March 20, 2003"
#define DRV_NAME "bonding"
#define DRV_DESCRIPTION "Ethernet Channel Bonding Driver"
......@@ -330,14 +368,30 @@ static char *primary= NULL;
static int max_bonds = BOND_DEFAULT_MAX_BONDS;
static int miimon = BOND_LINK_MON_INTERV;
static int mode = BOND_MODE_ROUNDROBIN;
static int use_carrier = 1;
static int bond_mode = BOND_MODE_ROUNDROBIN;
static int updelay = 0;
static int downdelay = 0;
#define BOND_MULTICAST_DISABLED 0
#define BOND_MULTICAST_ACTIVE 1
#define BOND_MULTICAST_ALL 2
static int multicast = BOND_MULTICAST_ALL;
static char *mode = NULL;
static struct bond_parm_tbl bond_mode_tbl[] = {
{ "balance-rr", BOND_MODE_ROUNDROBIN},
{ "active-backup", BOND_MODE_ACTIVEBACKUP},
{ "balance-xor", BOND_MODE_XOR},
{ "broadcast", BOND_MODE_BROADCAST},
{ NULL, -1},
};
static int multicast_mode = BOND_MULTICAST_ALL;
static char *multicast = NULL;
static struct bond_parm_tbl bond_mc_tbl[] = {
{ "disabled", BOND_MULTICAST_DISABLED},
{ "active", BOND_MULTICAST_ACTIVE},
{ "all", BOND_MULTICAST_ALL},
{ NULL, -1},
};
static int first_pass = 1;
static struct bonding *these_bonds = NULL;
......@@ -347,25 +401,23 @@ MODULE_PARM(max_bonds, "i");
MODULE_PARM_DESC(max_bonds, "Max number of bonded devices");
MODULE_PARM(miimon, "i");
MODULE_PARM_DESC(miimon, "Link check interval in milliseconds");
MODULE_PARM(mode, "i");
MODULE_PARM(use_carrier, "i");
MODULE_PARM_DESC(use_carrier, "Use netif_carrier_ok (vs MII ioctls) in miimon; 09 for off, 1 for on (default)");
MODULE_PARM(mode, "s");
MODULE_PARM_DESC(mode, "Mode of operation : 0 for round robin, 1 for active-backup, 2 for xor");
MODULE_PARM(arp_interval, "i");
MODULE_PARM_DESC(arp_interval, "arp interval in milliseconds");
MODULE_PARM(arp_ip_target, "1-" __MODULE_STRING(MAX_ARP_IP_TARGETS) "s");
MODULE_PARM_DESC(arp_ip_target, "arp targets in n.n.n.n form");
MODULE_PARM_DESC(mode, "Mode of operation : 0 for round robin, 1 for active-backup, 2 for xor");
MODULE_PARM(updelay, "i");
MODULE_PARM_DESC(updelay, "Delay before considering link up, in milliseconds");
MODULE_PARM(downdelay, "i");
MODULE_PARM_DESC(downdelay, "Delay before considering link down, in milliseconds");
MODULE_PARM(primary, "s");
MODULE_PARM_DESC(primary, "Primary network device to use");
MODULE_PARM(multicast, "i");
MODULE_PARM(multicast, "s");
MODULE_PARM_DESC(multicast, "Mode for multicast support : 0 for none, 1 for active slave, 2 for all slaves (default)");
extern void arp_send( int type, int ptype, u32 dest_ip, struct net_device *dev,
u32 src_ip, unsigned char *dest_hw, unsigned char *src_hw,
unsigned char *target_hw);
static int bond_xmit_roundrobin(struct sk_buff *skb, struct net_device *dev);
static int bond_xmit_xor(struct sk_buff *skb, struct net_device *dev);
static int bond_xmit_activebackup(struct sk_buff *skb, struct net_device *dev);
......@@ -409,7 +461,8 @@ static int bond_get_info(char *buf, char **start, off_t offset, int length);
static void arp_send_all(slave_t *slave)
{
int i;
for ( i=0; (i<MAX_ARP_IP_TARGETS) && arp_target[i]; i++) {
for (i = 0; (i<MAX_ARP_IP_TARGETS) && arp_target[i]; i++) {
arp_send(ARPOP_REQUEST, ETH_P_ARP, arp_target[i], slave->dev,
my_ip, arp_target_hw_addr, slave->dev->dev_addr,
arp_target_hw_addr);
......@@ -417,9 +470,10 @@ static void arp_send_all(slave_t *slave)
}
static const char *bond_mode(void)
static const char *
bond_mode_name(void)
{
switch (mode) {
switch (bond_mode) {
case BOND_MODE_ROUNDROBIN :
return "load balancing (round-robin)";
case BOND_MODE_ACTIVEBACKUP :
......@@ -433,9 +487,10 @@ static const char *bond_mode(void)
}
}
static const char *multicast_mode(void)
static const char *
multicast_mode_name(void)
{
switch(multicast) {
switch(multicast_mode) {
case BOND_MULTICAST_DISABLED :
return "disabled";
case BOND_MULTICAST_ACTIVE :
......@@ -464,6 +519,29 @@ static void bond_set_slave_active_flags(slave_t *slave)
slave->dev->flags &= ~IFF_NOARP;
}
/*
* This function counts and verifies the the number of attached
* slaves, checking the count against the expected value (given that incr
* is either 1 or -1, for add or removal of a slave). Only
* bond_xmit_xor() uses the slave_cnt value, but this is still a good
* consistency check.
*/
static inline void
update_slave_cnt(bonding_t *bond, int incr)
{
slave_t *slave = NULL;
int expect = bond->slave_cnt + incr;
bond->slave_cnt = 0;
for (slave = bond->prev; slave != (slave_t*)bond;
slave = slave->prev) {
bond->slave_cnt++;
}
if (expect != bond->slave_cnt)
BUG();
}
/*
* This function detaches the slave <slave> from the list <bond>.
* WARNING: no check is made to verify if the slave effectively
......@@ -471,8 +549,11 @@ static void bond_set_slave_active_flags(slave_t *slave)
* Nothing is freed on return, structures are just unchained.
* If the bond->current_slave pointer was pointing to <slave>,
* it's replaced with slave->next, or <bond> if not applicable.
*
* bond->lock held by caller.
*/
static slave_t *bond_detach_slave(bonding_t *bond, slave_t *slave)
static slave_t *
bond_detach_slave(bonding_t *bond, slave_t *slave)
{
if ((bond == NULL) || (slave == NULL) ||
((void *)bond == (void *)slave)) {
......@@ -499,8 +580,7 @@ static slave_t *bond_detach_slave(bonding_t *bond, slave_t *slave)
}
write_unlock(&bond->ptrlock);
}
}
else {
} else {
slave->prev->next = slave->next;
if (bond->prev == slave) { /* is this slave the last one ? */
bond->prev = slave->prev;
......@@ -515,9 +595,44 @@ static slave_t *bond_detach_slave(bonding_t *bond, slave_t *slave)
write_unlock(&bond->ptrlock);
}
update_slave_cnt(bond, -1);
return slave;
}
static void
bond_attach_slave(struct bonding *bond, struct slave *new_slave)
{
/*
* queue to the end of the slaves list, make the first element its
* successor, the last one its predecessor, and make it the bond's
* predecessor.
*
* Just to clarify, so future bonding driver hackers don't go through
* the same confusion stage I did trying to figure this out, the
* slaves are stored in a double linked circular list, sortof.
* In the ->next direction, the last slave points to the first slave,
* bypassing bond; only the slaves are in the ->next direction.
* In the ->prev direction, however, the first slave points to bond
* and bond points to the last slave.
*
* It looks like a circle with a little bubble hanging off one side
* in the ->prev direction only.
*
* When going through the list once, its best to start at bond->prev
* and go in the ->prev direction, testing for bond. Doing this
* in the ->next direction doesn't work. Trust me, I know this now.
* :) -mts 2002.03.14
*/
new_slave->prev = bond->prev;
new_slave->prev->next = new_slave;
bond->prev = new_slave;
new_slave->next = bond->next;
update_slave_cnt(bond, 1);
}
/*
* Less bad way to call ioctl from within the kernel; this needs to be
* done some other way to get the call out of interrupt context.
......@@ -534,17 +649,31 @@ static slave_t *bond_detach_slave(bonding_t *bond, slave_t *slave)
/*
* if <dev> supports MII link status reporting, check its link status.
*
* We either do MII/ETHTOOL ioctls, or check netif_carrier_ok(),
* depening upon the setting of the use_carrier parameter.
*
* Return either BMSR_LSTATUS, meaning that the link is up (or we
* can't tell and just pretend it is), or 0, meaning that the link is
* down.
*/
static u16 bond_check_dev_link(struct net_device *dev)
*
* If reporting is non-zero, instead of faking link up, return -1 if
* both ETHTOOL and MII ioctls fail (meaning the device does not
* support them). If use_carrier is set, return whatever it says.
* It'd be nice if there was a good way to tell if a driver supports
* netif_carrier, but there really isn't.
*/
static int
bond_check_dev_link(struct net_device *dev, int reporting)
{
static int (* ioctl)(struct net_device *, struct ifreq *, int);
struct ifreq ifr;
struct mii_ioctl_data *mii;
struct ethtool_value etool;
if (use_carrier) {
return netif_carrier_ok(dev) ? BMSR_LSTATUS : 0;
}
ioctl = dev->do_ioctl;
if (ioctl) {
/* TODO: set pointer to correct ioctl on a per team member */
......@@ -576,18 +705,24 @@ static u16 bond_check_dev_link(struct net_device *dev)
if (IOCTL(dev, &ifr, SIOCETHTOOL) == 0) {
if (etool.data == 1) {
return BMSR_LSTATUS;
}
else {
} else {
#ifdef BONDING_DEBUG
printk(KERN_INFO
":: SIOCETHTOOL shows failure \n");
":: SIOCETHTOOL shows link down \n");
#endif
return(0);
return 0;
}
}
}
return BMSR_LSTATUS; /* spoof link up ( we can't check it) */
/*
* If reporting, report that either there's no dev->do_ioctl,
* or both SIOCGMIIREG and SIOCETHTOOL failed (meaning that we
* cannot report link status). If not reporting, pretend
* we're ok.
*/
return reporting ? -1 : BMSR_LSTATUS;
}
static u16 bond_check_mii_link(bonding_t *bond)
......@@ -622,7 +757,7 @@ static int bond_open(struct net_device *dev)
init_timer(arp_timer);
arp_timer->expires = jiffies + (arp_interval * HZ / 1000);
arp_timer->data = (unsigned long)dev;
if (mode == BOND_MODE_ACTIVEBACKUP) {
if (bond_mode == BOND_MODE_ACTIVEBACKUP) {
arp_timer->function = (void *)&activebackup_arp_monitor;
} else {
arp_timer->function = (void *)&loadbalance_arp_monitor;
......@@ -692,7 +827,7 @@ static void bond_mc_list_destroy(struct bonding *bond)
static void bond_mc_add(bonding_t *bond, void *addr, int alen)
{
slave_t *slave;
switch (multicast) {
switch (multicast_mode) {
case BOND_MULTICAST_ACTIVE :
/* write lock already acquired */
if (bond->current_slave != NULL)
......@@ -713,7 +848,7 @@ static void bond_mc_add(bonding_t *bond, void *addr, int alen)
static void bond_mc_delete(bonding_t *bond, void *addr, int alen)
{
slave_t *slave;
switch (multicast) {
switch (multicast_mode) {
case BOND_MULTICAST_ACTIVE :
/* write lock already acquired */
if (bond->current_slave != NULL)
......@@ -769,7 +904,7 @@ static inline int dmi_same(struct dev_mc_list *dmi1, struct dev_mc_list *dmi2)
static void bond_set_promiscuity(bonding_t *bond, int inc)
{
slave_t *slave;
switch (multicast) {
switch (multicast_mode) {
case BOND_MULTICAST_ACTIVE :
/* write lock already acquired */
if (bond->current_slave != NULL)
......@@ -790,7 +925,7 @@ static void bond_set_promiscuity(bonding_t *bond, int inc)
static void bond_set_allmulti(bonding_t *bond, int inc)
{
slave_t *slave;
switch (multicast) {
switch (multicast_mode) {
case BOND_MULTICAST_ACTIVE :
/* write lock already acquired */
if (bond->current_slave != NULL)
......@@ -827,7 +962,7 @@ static void set_multicast_list(struct net_device *master)
struct dev_mc_list *dmi;
unsigned long flags = 0;
if (multicast == BOND_MULTICAST_DISABLED)
if (multicast_mode == BOND_MULTICAST_DISABLED)
return;
/*
* Lock the private data for the master
......@@ -865,7 +1000,7 @@ static void set_multicast_list(struct net_device *master)
/* save master's multicast list */
bond_mc_list_destroy (bond);
bond_mc_list_copy (master->mc_list, bond, GFP_KERNEL);
bond_mc_list_copy (master->mc_list, bond, GFP_ATOMIC);
write_unlock_irqrestore(&bond->lock, flags);
}
......@@ -878,7 +1013,7 @@ static void bond_mc_update(bonding_t *bond, slave_t *new, slave_t *old)
{
struct dev_mc_list *dmi;
switch(multicast) {
switch(multicast_mode) {
case BOND_MULTICAST_ACTIVE :
if (bond->device->flags & IFF_PROMISC) {
if (old != NULL && new != old)
......@@ -907,20 +1042,6 @@ static void bond_mc_update(bonding_t *bond, slave_t *new, slave_t *old)
}
}
/*
* This function counts the number of attached
* slaves for use by bond_xmit_xor.
*/
static void update_slave_cnt(bonding_t *bond)
{
slave_t *slave = NULL;
bond->slave_cnt = 0;
for (slave = bond->prev; slave != (slave_t*)bond; slave = slave->prev) {
bond->slave_cnt++;
}
}
/* enslave device <slave> to bond device <master> */
static int bond_enslave(struct net_device *master_dev,
struct net_device *slave_dev)
......@@ -934,10 +1055,7 @@ static int bond_enslave(struct net_device *master_dev,
struct dev_mc_list *dmi;
struct in_ifaddr **ifap;
struct in_ifaddr *ifa;
static int (* ioctl)(struct net_device *, struct ifreq *, int);
struct ifreq ifr;
struct ethtool_value etool;
int link_reporting = 0;
int link_reporting;
if (master_dev == NULL || slave_dev == NULL) {
return -ENODEV;
......@@ -949,14 +1067,12 @@ static int bond_enslave(struct net_device *master_dev,
"Warning : no link monitoring support for %s\n",
slave_dev->name);
}
write_lock_irqsave(&bond->lock, flags);
/* not running. */
if ((slave_dev->flags & IFF_UP) != IFF_UP) {
#ifdef BONDING_DEBUG
printk(KERN_CRIT "Error, slave_dev is not running\n");
#endif
write_unlock_irqrestore(&bond->lock, flags);
return -EINVAL;
}
......@@ -965,12 +1081,10 @@ static int bond_enslave(struct net_device *master_dev,
#ifdef BONDING_DEBUG
printk(KERN_CRIT "Error, Device was already enslaved\n");
#endif
write_unlock_irqrestore(&bond->lock, flags);
return -EBUSY;
}
if ((new_slave = kmalloc(sizeof(slave_t), GFP_KERNEL)) == NULL) {
write_unlock_irqrestore(&bond->lock, flags);
if ((new_slave = kmalloc(sizeof(slave_t), GFP_ATOMIC)) == NULL) {
return -ENOMEM;
}
memset(new_slave, 0, sizeof(slave_t));
......@@ -983,14 +1097,12 @@ static int bond_enslave(struct net_device *master_dev,
#ifdef BONDING_DEBUG
printk(KERN_CRIT "Error %d calling netdev_set_master\n", err);
#endif
kfree(new_slave);
write_unlock_irqrestore(&bond->lock, flags);
return err;
goto err_free;
}
new_slave->dev = slave_dev;
if (multicast == BOND_MULTICAST_ALL) {
if (multicast_mode == BOND_MULTICAST_ALL) {
/* set promiscuity level to new slave */
if (master_dev->flags & IFF_PROMISC)
dev_set_promiscuity(slave_dev, 1);
......@@ -1004,64 +1116,24 @@ static int bond_enslave(struct net_device *master_dev,
dev_mc_add (slave_dev, dmi->dmi_addr, dmi->dmi_addrlen, 0);
}
/*
* queue to the end of the slaves list, make the first element its
* successor, the last one its predecessor, and make it the bond's
* predecessor.
*
* Just to clarify, so future bonding driver hackers don't go through
* the same confusion stage I did trying to figure this out, the
* slaves are stored in a double linked circular list, sortof.
* In the ->next direction, the last slave points to the first slave,
* bypassing bond; only the slaves are in the ->next direction.
* In the ->prev direction, however, the first slave points to bond
* and bond points to the last slave.
*
* It looks like a circle with a little bubble hanging off one side
* in the ->prev direction only.
*
* When going through the list once, its best to start at bond->prev
* and go in the ->prev direction, testing for bond. Doing this
* in the ->next direction doesn't work. Trust me, I know this now.
* :) -mts 2002.03.14
*/
new_slave->prev = bond->prev;
new_slave->prev->next = new_slave;
bond->prev = new_slave;
new_slave->next = bond->next;
write_lock_irqsave(&bond->lock, flags);
bond_attach_slave(bond, new_slave);
new_slave->delay = 0;
new_slave->link_failure_count = 0;
if (miimon > 0) {
/* if the network driver for the slave does not support
* ETHTOOL/MII link status reporting, warn the user of this
*/
if ((ioctl = slave_dev->do_ioctl) != NULL) {
etool.cmd = ETHTOOL_GLINK;
ifr.ifr_data = (char*)&etool;
if (IOCTL(slave_dev, &ifr, SIOCETHTOOL) == 0) {
link_reporting = 1;
}
else {
if (IOCTL(slave_dev, &ifr, SIOCGMIIPHY) == 0) {
/* Yes, the mii is overlaid on the
* ifreq.ifr_ifru
*/
((struct mii_ioctl_data*)
(&ifr.ifr_data))->reg_num = 1;
if (IOCTL(slave_dev, &ifr, SIOCGMIIREG)
== 0) {
link_reporting = 1;
}
}
}
}
if (miimon > 0 && !use_carrier) {
link_reporting = bond_check_dev_link(slave_dev, 1);
if ((link_reporting == 0) && (arp_interval == 0)) {
/* miimon is set but a bonded network driver does
* not support ETHTOOL/MII and arp_interval is
* not set
if ((link_reporting == -1) && (arp_interval == 0)) {
/*
* miimon is set but a bonded network driver
* does not support ETHTOOL/MII and
* arp_interval is not set. Note: if
* use_carrier is enabled, we will never go
* here (because netif_carrier is always
* supported); thus, we don't need to change
* the messages for netif_carrier.
*/
printk(KERN_ERR
"bond_enslave(): MII and ETHTOOL support not "
......@@ -1070,8 +1142,7 @@ static int bond_enslave(struct net_device *master_dev,
"not specified, thus bonding will not detect "
"link failures! see bonding.txt for details.\n",
slave_dev->name);
}
else if (link_reporting == 0) {
} else if (link_reporting == -1) {
/* unable get link status using mii/ethtool */
printk(KERN_WARNING
"bond_enslave: can't get link status from "
......@@ -1085,7 +1156,7 @@ static int bond_enslave(struct net_device *master_dev,
/* check for initial state */
if ((miimon <= 0) ||
(bond_check_dev_link(slave_dev) == BMSR_LSTATUS)) {
(bond_check_dev_link(slave_dev, 0) == BMSR_LSTATUS)) {
#ifdef BONDING_DEBUG
printk(KERN_CRIT "Initial state of slave_dev is BOND_LINK_UP\n");
#endif
......@@ -1106,7 +1177,7 @@ static int bond_enslave(struct net_device *master_dev,
* since we guarantee that current_slave always point to the last
* usable interface, we just have to verify this interface's flag.
*/
if (mode == BOND_MODE_ACTIVEBACKUP) {
if (bond_mode == BOND_MODE_ACTIVEBACKUP) {
if (((bond->current_slave == NULL)
|| (bond->current_slave->dev->flags & IFF_NOARP))
&& (new_slave->link == BOND_LINK_UP)) {
......@@ -1145,8 +1216,6 @@ static int bond_enslave(struct net_device *master_dev,
bond->current_slave = new_slave;
}
update_slave_cnt(bond);
write_unlock_irqrestore(&bond->lock, flags);
/*
......@@ -1187,7 +1256,11 @@ static int bond_enslave(struct net_device *master_dev,
new_slave->state == BOND_STATE_ACTIVE ? "n active" : " backup",
new_slave->link == BOND_LINK_UP ? "n up" : " down");
/* enslave is successful */
return 0;
err_free:
kfree(new_slave);
return err;
}
/*
......@@ -1276,13 +1349,13 @@ slave_t *change_active_interface(bonding_t *bond)
} else {
printk (" but could not find any %s interface.\n",
(mode == BOND_MODE_ACTIVEBACKUP) ? "backup":"other");
(bond_mode == BOND_MODE_ACTIVEBACKUP) ? "backup":"other");
write_lock(&bond->ptrlock);
bond->current_slave = (slave_t *)NULL;
write_unlock(&bond->ptrlock);
return NULL; /* still no slave, return NULL */
}
} else if (mode == BOND_MODE_ACTIVEBACKUP) {
} else if (bond_mode == BOND_MODE_ACTIVEBACKUP) {
/* make sure oldslave doesn't send arps - this could
* cause a ping-pong effect between interfaces since they
* would be able to tx arps - in active backup only one
......@@ -1311,7 +1384,7 @@ slave_t *change_active_interface(bonding_t *bond)
if (IS_UP(newslave->dev)) {
if (newslave->link == BOND_LINK_UP) {
/* this one is immediately usable */
if (mode == BOND_MODE_ACTIVEBACKUP) {
if (bond_mode == BOND_MODE_ACTIVEBACKUP) {
bond_set_slave_active_flags(newslave);
bond_mc_update(bond, newslave, oldslave);
printk (" and making interface %s the active one.\n",
......@@ -1358,8 +1431,8 @@ slave_t *change_active_interface(bonding_t *bond)
return bestslave;
}
if ((mode == BOND_MODE_ACTIVEBACKUP) &&
(multicast == BOND_MULTICAST_ACTIVE) &&
if ((bond_mode == BOND_MODE_ACTIVEBACKUP) &&
(multicast_mode == BOND_MULTICAST_ACTIVE) &&
(oldslave != NULL)) {
/* flush bonds (master's) mc_list from oldslave since it wasn't
* updated (and deleted) above
......@@ -1374,7 +1447,7 @@ slave_t *change_active_interface(bonding_t *bond)
}
printk (" but could not find any %s interface.\n",
(mode == BOND_MODE_ACTIVEBACKUP) ? "backup":"other");
(bond_mode == BOND_MODE_ACTIVEBACKUP) ? "backup":"other");
/* absolutely nothing found. let's return NULL */
write_lock(&bond->ptrlock);
......@@ -1406,16 +1479,14 @@ static int bond_release(struct net_device *master, struct net_device *slave)
bond = (struct bonding *) master->priv;
write_lock_irqsave(&bond->lock, flags);
/* master already enslaved, or slave not enslaved,
or no slave for this master */
if ((master->flags & IFF_SLAVE) || !(slave->flags & IFF_SLAVE)) {
printk (KERN_DEBUG "%s: cannot release %s.\n", master->name, slave->name);
write_unlock_irqrestore(&bond->lock, flags);
return -EINVAL;
}
write_lock_irqsave(&bond->lock, flags);
bond->current_arp_slave = NULL;
our_slave = (slave_t *)bond;
old_current = bond->current_slave;
......@@ -1435,9 +1506,31 @@ static int bond_release(struct net_device *master, struct net_device *slave)
printk(".\n");
}
/* release the slave from its bond */
if (bond->current_slave == NULL) {
printk(KERN_INFO
"%s: now running without any active interface !\n",
master->name);
}
if (bond->primary_slave == our_slave) {
bond->primary_slave = NULL;
}
break;
}
}
write_unlock_irqrestore(&bond->lock, flags);
if (multicast == BOND_MULTICAST_ALL) {
if (our_slave == (slave_t *)bond) {
/* if we get here, it's because the device was not found */
printk (KERN_INFO "%s: %s not enslaved\n", master->name, slave->name);
return -EINVAL;
}
/* undo settings and restore original values */
if (multicast_mode == BOND_MULTICAST_ALL) {
/* flush master's mc_list from slave */
bond_mc_list_flush (slave, master);
......@@ -1464,30 +1557,10 @@ static int bond_release(struct net_device *master, struct net_device *slave)
}
bond_restore_slave_flags(our_slave);
kfree(our_slave);
if (bond->current_slave == NULL) {
printk(KERN_INFO
"%s: now running without any active interface !\n",
master->name);
}
update_slave_cnt(bond);
if (bond->primary_slave == our_slave) {
bond->primary_slave = NULL;
}
kfree(our_slave);
write_unlock_irqrestore(&bond->lock, flags);
return 0; /* deletion OK */
}
}
/* if we get here, it's because the device was not found */
write_unlock_irqrestore(&bond->lock, flags);
printk (KERN_INFO "%s: %s not enslaved\n", master->name, slave->name);
return -EINVAL;
}
/*
......@@ -1510,13 +1583,15 @@ static int bond_release_all(struct net_device *master)
bond = (struct bonding *) master->priv;
bond->current_arp_slave = NULL;
bond->current_slave = NULL;
bond->primary_slave = NULL;
while ((our_slave = bond->prev) != (slave_t *)bond) {
slave_dev = our_slave->dev;
bond->prev = our_slave->prev;
bond_detach_slave(bond, our_slave);
if (multicast == BOND_MULTICAST_ALL
|| (multicast == BOND_MULTICAST_ACTIVE
if (multicast_mode == BOND_MULTICAST_ALL
|| (multicast_mode == BOND_MULTICAST_ACTIVE
&& bond->current_slave == our_slave)) {
/* flush master's mc_list from slave */
......@@ -1533,6 +1608,10 @@ static int bond_release_all(struct net_device *master)
kfree(our_slave);
/*
* Can be safely called from inside the bond lock
* since traffic and timers have already stopped
*/
netdev_set_master(slave_dev, NULL);
/* only restore its RUNNING flag if monitoring set it down */
......@@ -1543,10 +1622,6 @@ static int bond_release_all(struct net_device *master)
dev_close(slave_dev);
}
bond->current_slave = NULL;
bond->next = (slave_t *)bond;
bond->slave_cnt = 0;
bond->primary_slave = NULL;
printk (KERN_INFO "%s: released all slaves\n", master->name);
return 0;
......@@ -1579,9 +1654,9 @@ static void bond_mii_monitor(struct net_device *master)
/* use updelay+1 to match an UP slave even when updelay is 0 */
int mindelay = updelay + 1;
struct net_device *dev = slave->dev;
u16 link_state;
int link_state;
link_state = bond_check_dev_link(dev);
link_state = bond_check_dev_link(dev, 0);
switch (slave->link) {
case BOND_LINK_UP: /* the link was up */
......@@ -1608,7 +1683,7 @@ static void bond_mii_monitor(struct net_device *master)
"%s, disabling it in %d ms.\n",
master->name,
IS_UP(dev)
? ((mode == BOND_MODE_ACTIVEBACKUP)
? ((bond_mode == BOND_MODE_ACTIVEBACKUP)
? ((slave == oldcurrent)
? "active " : "backup ")
: "")
......@@ -1628,7 +1703,7 @@ static void bond_mii_monitor(struct net_device *master)
slave->link = BOND_LINK_DOWN;
/* in active/backup mode, we must
completely disable this interface */
if (mode == BOND_MODE_ACTIVEBACKUP) {
if (bond_mode == BOND_MODE_ACTIVEBACKUP) {
bond_set_slave_inactive_flags(slave);
}
printk(KERN_INFO
......@@ -1708,14 +1783,12 @@ static void bond_mii_monitor(struct net_device *master)
slave->link = BOND_LINK_UP;
slave->jiffies = jiffies;
if ( (mode == BOND_MODE_ACTIVEBACKUP)
|| (slave != bond->primary_slave) ) {
/* prevent it from being the active one */
slave->state = BOND_STATE_BACKUP;
}
else {
if (bond_mode != BOND_MODE_ACTIVEBACKUP) {
/* make it immediately active */
slave->state = BOND_STATE_ACTIVE;
} else if (slave != bond->primary_slave) {
/* prevent it from being the active one */
slave->state = BOND_STATE_BACKUP;
}
printk(KERN_INFO
......@@ -1775,7 +1848,7 @@ static void bond_mii_monitor(struct net_device *master)
bestslave->jiffies = jiffies;
}
if (mode == BOND_MODE_ACTIVEBACKUP) {
if (bond_mode == BOND_MODE_ACTIVEBACKUP) {
bond_set_slave_active_flags(bestslave);
bond_mc_update(bond, bestslave, NULL);
} else {
......@@ -2171,19 +2244,6 @@ static void activebackup_arp_monitor(struct net_device *master)
read_unlock_irqrestore(&bond->lock, flags);
}
#define isdigit(c) (c >= '0' && c <= '9')
__inline static int atoi( char **s)
{
int i = 0;
while (isdigit(**s))
i = i*20 + *((*s)++) - '0';
return i;
}
#define isascii(c) (((unsigned char)(c))<=0x7f)
#define LF 0xA
#define isspace(c) (c==' ' || c==' '|| c==LF)
typedef uint32_t in_addr_t;
int
......@@ -2279,7 +2339,7 @@ static int bond_info_query(struct net_device *master, struct ifbond *info)
slave_t *slave;
unsigned long flags;
info->bond_mode = mode;
info->bond_mode = bond_mode;
info->num_slaves = 0;
info->miimon = miimon;
......@@ -2416,7 +2476,7 @@ static int bond_ioctl(struct net_device *master_dev, struct ifreq *ifr, int cmd)
break;
case BOND_CHANGE_ACTIVE_OLD:
case SIOCBONDCHANGEACTIVE:
if (mode == BOND_MODE_ACTIVEBACKUP) {
if (bond_mode == BOND_MODE_ACTIVEBACKUP) {
ret = bond_change_active(master_dev, slave_dev);
}
else {
......@@ -2567,15 +2627,8 @@ static int bond_xmit_xor(struct sk_buff *skb, struct net_device *dev)
slave = bond->prev;
/* we're at the root, get the first slave */
if ((slave == NULL) || (slave->dev == NULL)) {
/* no suitable interface, frame not sent */
dev_kfree_skb(skb);
read_unlock_irqrestore(&bond->lock, flags);
return 0;
}
if (bond->slave_cnt == 0) {
/* no slaves in the bond, frame not sent */
/* no suitable interface, frame not sent */
dev_kfree_skb(skb);
read_unlock_irqrestore(&bond->lock, flags);
return 0;
......@@ -2642,7 +2695,7 @@ static int bond_xmit_activebackup(struct sk_buff *skb, struct net_device *dev)
* receive packets to stay up, and the only ones they receive are
* broadcasts.
*/
if ( (mode != BOND_MODE_ACTIVEBACKUP) &&
if ( (bond_mode != BOND_MODE_ACTIVEBACKUP) &&
(arp_ip_count == 1) &&
(arp_interval > 0) && (arp_target_hw_addr == NULL) &&
(skb->protocol == __constant_htons(ETH_P_IP) ) ) {
......@@ -2743,9 +2796,10 @@ static int bond_get_info(char *buf, char **start, off_t offset, int length)
*/
link = bond_check_mii_link(bond);
len += sprintf(buf + len, "Bonding Mode: %s\n", bond_mode());
len += sprintf(buf + len, "Bonding Mode: %s\n",
bond_mode_name());
if (mode == BOND_MODE_ACTIVEBACKUP) {
if (bond_mode == BOND_MODE_ACTIVEBACKUP) {
read_lock_irqsave(&bond->lock, flags);
read_lock(&bond->ptrlock);
if (bond->current_slave != NULL) {
......@@ -2766,7 +2820,8 @@ static int bond_get_info(char *buf, char **start, off_t offset, int length)
updelay * miimon);
len += sprintf(buf + len, "Down Delay (ms): %d\n",
downdelay * miimon);
len += sprintf(buf + len, "Multicast Mode: %s\n", multicast_mode());
len += sprintf(buf + len, "Multicast Mode: %s\n",
multicast_mode_name());
read_lock_irqsave(&bond->lock, flags);
for (slave = bond->prev; slave != (slave_t *)bond;
......@@ -2884,16 +2939,21 @@ static int __init bond_init(struct net_device *dev)
dev->priv = bond;
/* Initialize the device structure. */
if (mode == BOND_MODE_ACTIVEBACKUP) {
switch (bond_mode) {
case BOND_MODE_ACTIVEBACKUP:
dev->hard_start_xmit = bond_xmit_activebackup;
} else if (mode == BOND_MODE_ROUNDROBIN) {
break;
case BOND_MODE_ROUNDROBIN:
dev->hard_start_xmit = bond_xmit_roundrobin;
} else if (mode == BOND_MODE_XOR) {
break;
case BOND_MODE_XOR:
dev->hard_start_xmit = bond_xmit_xor;
} else if (mode == BOND_MODE_BROADCAST) {
break;
case BOND_MODE_BROADCAST:
dev->hard_start_xmit = bond_xmit_broadcast;
} else {
printk(KERN_ERR "Unknown bonding mode %d\n", mode);
break;
default:
printk(KERN_ERR "Unknown bonding mode %d\n", bond_mode);
kfree(bond->stats);
kfree(bond);
return -EINVAL;
......@@ -2926,7 +2986,7 @@ static int __init bond_init(struct net_device *dev)
} else {
printk("out MII link monitoring");
}
printk(", in %s mode.\n", bond_mode());
printk(", in %s mode.\n", bond_mode_name());
printk(KERN_INFO "%s registered with", dev->name);
if (arp_interval > 0) {
......@@ -2986,6 +3046,28 @@ static int __init bond_probe(struct net_device *dev)
}
*/
/*
* Convert string input module parms. Accept either the
* number of the mode or its string name.
*/
static inline int
bond_parse_parm(char *mode_arg, struct bond_parm_tbl *tbl)
{
int i;
for (i = 0; tbl[i].modename != NULL; i++) {
if ((isdigit(*mode_arg) &&
tbl[i].mode == simple_strtol(mode_arg, NULL, 0)) ||
(0 == strncmp(mode_arg, tbl[i].modename,
strlen(tbl[i].modename)))) {
return tbl[i].mode;
}
}
return -1;
}
static int __init bonding_init(void)
{
int no;
......@@ -2996,6 +3078,29 @@ static int __init bonding_init(void)
printk(KERN_INFO "%s", version);
/*
* Convert string parameters.
*/
if (mode) {
bond_mode = bond_parse_parm(mode, bond_mode_tbl);
if (bond_mode == -1) {
printk(KERN_WARNING
"bonding_init(): Invalid bonding mode \"%s\"\n",
mode == NULL ? "NULL" : mode);
return -EINVAL;
}
}
if (multicast) {
multicast_mode = bond_parse_parm(multicast, bond_mc_tbl);
if (multicast_mode == -1) {
printk(KERN_WARNING
"bonding_init(): Invalid multicast mode \"%s\"\n",
multicast == NULL ? "NULL" : multicast);
return -EINVAL;
}
}
if (max_bonds < 1 || max_bonds > INT_MAX) {
printk(KERN_WARNING
"bonding_init(): max_bonds (%d) not in range %d-%d, "
......@@ -3125,28 +3230,18 @@ static int __init bonding_init(void)
"link failures! see bonding.txt for details.\n");
}
if ((primary != NULL) && (mode != BOND_MODE_ACTIVEBACKUP)){
if ((primary != NULL) && (bond_mode != BOND_MODE_ACTIVEBACKUP)){
/* currently, using a primary only makes sence
* in active backup mode
*/
printk(KERN_WARNING
"bonding_init(): %s primary device specified but has "
" no effect in %s mode\n",
primary, bond_mode());
primary, bond_mode_name());
primary = NULL;
}
if (multicast != BOND_MULTICAST_DISABLED &&
multicast != BOND_MULTICAST_ACTIVE &&
multicast != BOND_MULTICAST_ALL) {
printk(KERN_WARNING
"bonding_init(): unknown multicast module "
"parameter (%d), multicast reset to %d\n",
multicast, BOND_MULTICAST_ALL);
multicast = BOND_MULTICAST_ALL;
}
for (no = 0; no < max_bonds; no++) {
dev_bond->init = bond_init;
......
include/linux/if_bonding.h
View file @ 318103c4
......@@ -54,6 +54,15 @@
#define BOND_DEFAULT_MAX_BONDS 1 /* Default maximum number of devices to support */
#define BOND_MULTICAST_DISABLED 0
#define BOND_MULTICAST_ACTIVE 1
#define BOND_MULTICAST_ALL 2
struct bond_parm_tbl {
char *modename;
int mode;
};
typedef struct ifbond {
__s32 bond_mode;
__s32 num_slaves;
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7